[YARN-1894] RM shutdown due to java.net.UnknownHostException

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Duplicate
• Fix Version/s: 2.4.0
• Component/s: resourcemanager
• Assignee: Mohammad Kamrul Islam
• Reporter: Mohammad Kamrul Islam
• Created: 29/Mar/14 19:59
• Updated: 10/Apr/14 13:11
• Resolved: 30/Mar/14 02:41

Description

Background:
----------------
I started Hadoop 2.3 on my Mac in my office network and submitted few jobs successfully. When i went to my home (new network), I submitted another job and it abruptly pulled down the RM service.

Error in RM log:

2014-03-29 12:28:56,754 INFO org.apache.hadoop.yarn.server.resourcemanager.security.RMDelegationTokenSecretManager: storing RMDelegation token with sequence number: 3
2014-03-29 12:28:57,256 FATAL org.apache.hadoop.yarn.server.resourcemanager.ResourceManager: Error in handling event type NODE_UPDATE to the scheduler
java.lang.IllegalArgumentException: java.net.UnknownHostException: mislam-mn.<MY.OOFICE.DOMAIN>
        at org.apache.hadoop.security.SecurityUtil.buildTokenService(SecurityUtil.java:377)
        at org.apache.hadoop.yarn.server.utils.BuilderUtils.newContainerToken(BuilderUtils.java:247)
        at org.apache.hadoop.yarn.server.resourcemanager.security.RMContainerTokenSecretManager.createContainerToken(RMContainerTokenSecretManager.java:195)
        at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.createContainerToken(LeafQueue.java:1294)
        at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.assignContainer(LeafQueue.java:1342)
        at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.assignOffSwitchContainers(LeafQueue.java:1208)
        at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.assignContainersOnNode(LeafQueue.java:1167)
        at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.assignContainers(LeafQueue.java:868)
        at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.ParentQueue.assignContainersToChildQueues(ParentQueue.java:642)
        at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.ParentQueue.assignContainers(ParentQueue.java:556)
        at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler.nodeUpdate(CapacityScheduler.java:696)
        at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler.handle(CapacityScheduler.java:740)
        at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler.handle(CapacityScheduler.java:88)
        at org.apache.hadoop.yarn.server.resourcemanager.ResourceManager$SchedulerEventDispatcher$EventProcessor.run(ResourceManager.java:543)
        at java.lang.Thread.run(Thread.java:695)
Caused by: java.net.UnknownHostException: mislam-mn.linkedin.biz
        ... 15 more
2014-03-29 12:28:57,259 INFO org.apache.hadoop.yarn.server.resourcemanager.ResourceManager: Exiting, bbye..
2014-03-29 12:28:57,297 INFO org.mortbay.log: Stopped SelectChannelConnector@0.0.0.0:8088
2014-03-29 12:28:57,401 INFO org.apache.hadoop.ipc.Server: Stopping server on 8032
2014-03-29 12:28:57,473 INFO org.apache.hadoop.ipc.Server: Stopping server on 8033
.....

Proposal:
---------------
I believe the root cause : I moved my machine from one network to another with the same RM service.

My point is: Whatever the cause, RM is a long running core-service and it should not exit this way. An appropriate error message should be sufficient.

If there is an consensus (or no disagreement), I can work for a patch.

Issue Links

• is duplicated by: YARN-713 ResourceManager can exit unexpectedly if DNS is unavailable